Publishable Final Activity Report

Transcription

1 SIXTH FRAMEWORK PROGRAMME PRIORITY [4] [Aeronautics and Space] Integrated Tool for Simulation of Textile Composites Contract no.: Publishable Final Activity Report Period covered: from to Start date of project: Duration: 42 months Internal reference: E Revision: issue 1 Project coordinator name: Dr. Peter Middendorf Project coordinator organisation name: EADS Deutschland GmbH Page 1 of 36

2 Content 1 Original project data General project data Abstract Objectives Partners Project Work packages and work package interdependence Project organisation structure Main advantage of the approach Main project motivation Work performed and results WP1: Material Characterisation / DataTool WP2: 3M Mechanics WP3: Process Simulation WP4: External loading behaviour WP5: Integration WP6: Industrial validation / Test WP7: Design guidelines / Standardisation List of deliverables Degree to which the objectives where reached Achievements of the project related to the state of the art State of the art before project start Achievements of the project Dissemination and use Exploitable knowledge and its use Dissemination of knowledge References Journal papers Conference contributions Other publications Page 2 of 36

3 1 Original project data 1.1 General project data Title: Integrated Tool for Simulation of Textile Composites Acronym: Contract Nr.: Starting Date: 01/03/2005 Duration: 42 months Total Cost: EU Contribution: Web-site: The project website will be kept online beyond the end of the project, dated Coordinator: EADS Deutschland GmbH Innovation Works CTO/IW-SP München Contact: Dr. Peter Middendorf Tel: +49 (0) Fax: +49 (0) Project logo: Page 3 of 36

4 1.2 Abstract The overall aim of is to increase the usage of textile preforming for composites in Aeronautics applications. It is well-known from related projects that preforming offers a potential cost savings of 20-30% in materials and processing compared with prepreg technology. But a still missing prerequisite for taking advantage of this potential are adequate design and analysis methods and especially validated simulation tools. The technical approach of the project covers this aspect by development of an integrated solution, simulating the manufacturing and processing chain as well as the loading stage to get reliable results also for 3d fibre architectures. Focus is laid on braiding, weaving and stitching technologies including also Non Crimp Fabrics. The resulting types of preforms will be analysed on different approximation levels (micro / meso / macro modelling) to take inhomogeneous fibre distribution and waviness into account. To fulfil the objectives within a limited time (and cost) scale, the linking and integration of different stand-alone solutions in the tool chain is proposed thus creating an open flexible interface for fluent data exchange and communication. The alternative to develop new multi purpose software would be prohibitively costly. The main value is therefore gained for the user of textile composites. can set up a standard for testing, modelling and simulation with minimized interference and without data loss and by this reply to the market demands. Further impact of the enhanced simulation capabilities will be a distinct reduction of at least 20% in necessary testing effort as well as a lead time reduction of more than 15%. The proof of this concept will be performed for different application examples in Aeronautics and summarized in guidelines, design rules and educational tools. 1.3 Objectives Textile preforming of composites offers the potential of significant cost savings of 20-30% in comparison to prepreg tape layering due to cheaper materials (dry fibre rovings, yarns or fabrics), easier storage conditions and the possibility of automation of the preforming process. These advantages open new markets for textile preforming, especially in low-cost composite applications. To enable development engineers to make use of dry fibre textiles, reliable simulation tools and design principles are needed. But in contrast to conventional, unidirectional reinforced composites, textile reinforcement results in 3d fibre architectures so that standard analysis procedures like 2d rules of mixture, in-plane strength criteria and classical laminate theory are no longer valid. For textile preforming it is also important to consider the manufacturing of the textile preform since the shaping operation will cause fibre reorientations, distributions, frictional and prestressing forces which will have a strong influence on the resulting textile properties. The technical approach of is therefore a simulation along the process line with a virtual manufacturing chain incorporating the preform manufacturing, draping and impregnation process followed by the external loading of the finished component. The scientific objective of is to close the gap between missing knowledge and proved advantages of dry fibre textiles by development of an adequate integrated simulation tool for textile preforming technologies including braiding, advanced engineering textiles, weaving and stitching. Reliable simulation tools and design methods provide the enabling prerequisites for an increased use of these materials in Aerospace (and other) industries. Page 4 of 36

5 From the technical point of view, special focus will be laid on 3d reinforcement by the use of structural stitching to improve mechanical properties of composites in the thickness direction (damage tolerance +80%, fracture toughness +75%, weight specific energy absorption +75%). The mechanical behaviour will be analysed on three different approximation levels called 3M (micro / meso / macro) mechanics: on the microscale the different constituents are always modelled separately, on the mesoscale fibre and matrix properties are homogenized locally, on macro level the micro or mesoscale models are homogenized in a coarser way to lower the computational effort. As there are already stand-alone solutions for several parts of the simulation in use, the approach of is mainly the linking and integration of these tools to ensure a fluid interaction and data interchange with minimum friction and without critical data loss. Furthermore the user interference should be minimised. This approach will enable a flexible and adaptable solution, which may be extended to include alternative technologies such as Liquid Composite Moulding and impregnation simulation. The proof for this integration concept will be performed for different application fields of textile preformed composites in Aerospace: typical stiffened skin sections, integral joining technologies and a braided propeller fan. The evaluation also includes the interface and the related flow of data as the quality of results in comparison with tests. The achievement of reliable simulation data will lead to a further 20% saving of testing effort and to a lead time reduction of approximately 15% due to a decrease of product development iteration loops. Greater optimisation of structural parts via validated simulation tools will be possible. In parallel to the development of the integrated simulation tool the second aspect of the project is to build up physical understanding of textile preformed composites behaviour to increase their usage. Therefore design rules for the use of dry fibre textiles should be extracted and made easily available for the design engineer in a guideline. In addition, an educational tool based on e-learning concepts will be developed and disseminated. By achieving the above-mentioned objectives, can provide the basis of a standard for the design, analysis and testing of textile preformed composites in Europe. 1.4 Partners The consortium comprises three aerospace companies, one SME, one software supplier, two research organisations and six universities to reflect applications, scientific and research partners of the project. All of the partners are experienced in either textile composites preforming, simulation or both of them and have in many cases already worked in previous national or European projects. A summary of participating companies and institutes is given below in table 1. Table 1: participants Partic. Role 1 Partic. no. Participant name Participant short name Country Date enter project 2 Date exit project CO 1 EADS Germany EADS-G D Month 1 Month 42 1 CO = Coordinator, CR = Contractor 2 Normally insert month 1 (start of project) and month n (end of project) These columns are needed for possible later contract revisions caused by joining/leaving participants Page 5 of 36

6 CR 2 Alenia Aeronautica S.p.A. ALA IT Month 1 Month 42 CR 3 Cranfield University CRAN GB Month 1 Month 42 CR 4 Dassault Aviation DAS F Month 1 Month 42 CR 5 German Aerospace Center DLR D Month 1 Month 42 CR 6 EADS France Innovation Works EADS F IW F Month 1 Month 42 CR 7 ESI Software ESI F Month 1 Month 42 CR 8 University of Stuttgart IFB D Month 1 Month 42 CR 9 University of Aachen ITA D Month 1 Month 42 CR 10 University of Leuven KUL B Month 1 Month 42 CR 11 INSA University Lyon LAMCOS F Month 1 Month 42 CR 12 SISPRA SISPRA E Month 1 Month 42 CR 13 University of Zaragoza ZARA E Month 1 Month 42 * CO = Coordinator CR = Contractor ** Normally insert month 1 (start of project) and month n (end of project) These columns are needed for possible later contract revisions caused by joining/leaving participants 1.5 Project Work packages and work package interdependence Covering the whole development chain, the project starts with WP1 Material Characterisation / DataTool where the requirements for the DataTool used in other parts of is researched. Materials, used in the validation examples, are analysed and their properties needed for the validation examples are stored in the DataTool. In parallel, WP2 3M Mechanics provides models compatible with and ready to be integrated in structural analysis tools developed on Macro level. It also contributes to the description of the textile geometry, to the development of textile process models and stiffness/failure/impact models. WP3 Process Simulation develops a `virtual manufacturing chain for the production of textile reinforced plastics. WP4 External loading behaviour investigates the complete structure of textile reinforced plastics and aims to integrate macroscopic structural deformation, stress and failure modelling into. WP5 Integration takes care of the interaction between these different WP solutions. The interaction is defined properly for each component and described in such a way, that data interchanging is made with minimum friction and without critical data loss. User interference is also minimized by the use of a Graphical User Interface (GUI). Using three different test cases, WP6 Industrial validation / Test validates the whole simulation chain by a comparison between the manufacturing and the process simulation as well as between the mechanical testing and external loading analysis. Finally, WP7 Design guidelines / Standardisation provides general design guidelines, design diagrams and rule-of-thumb design procedures to allow textile engineers as well as composite engineers to take basic decisions at a very early phase of the development and to allow a rough estimation concerning the performance of a structure. The workflow and interaction of work packages within the project is depicted in the chart below (figure 1). Page 6 of 36

7 WP1: Material Characterisation / Datatool Material Characterisation Datatool WP2: 3M Mechanics Micro Meso Macro I/O Interface Tool Interface GUI WP5: Integration / Specification WP3: Process Simulation Textile Process Draping Infiltration WP4: External Loading Behaviour Static Impact Failure Strength & Crash WP6: Industrial Validation / Test WP7: Design Guidelines / Standardisation Design Guidelines Standardisation WP0: Project Management Figure 1: Work flow diagram 1.6 Project organisation structure The project structure was organised according to ISO standard to ensure fluent communication between partners. A site manager which is responsible for the project contributions was appointed for each partner. A work package leader was responsible for the work in each work package. On a regular base (each six month) the technical am management status of the project was discussed during a review meeting and have been reported in technical reports. A graphical overview of the project organisation structure is shown in figure 2. Page 7 of 36

8 PMC PM Exploitation manager Site Managers Contractor 1 Contractor 2 Contractor WP leaders WPL1 WPL2 WPL3 WPL4 WPL5 WPL6 WPL7 Figure 2: Overview of project management structure 1.7 Main advantage of the approach The advantages of the approach against the up to now standard approach are listed in table 2. Table 2: comparison between standard and approach. Standard approach A significant amount of experimental testing is required to characterise new materials The selection of the textile reinforcement is limited to the existing products on the market Uniform permeability s re used to perform an infiltration simulation. For the draping process no or at maximum kinematic draping algorithms are used which doe not take into account the mechanical properties of the textile Only the average behaviour of the textile reinforced composite is know from the experimentally investigations Stand alone solution are used to perform different steps in the design process of a (textile) composite part. Often problems arise when transferring the results between different simulation steps No universal methods and standard exist for textile reinforced composites due to the complexity and variability of these materials. approach Many of the material properties can be simulated and validated by a limited amount of experimental work. Engineers can virtually design optimised textile architectures for a specific application A mechanically based draping algorithm predicts the textile deformation during the draping process. Based on these results and a detailed geometrical description of the deformed textile a local permeability is predicted for an infiltration simulation. The multi-level modelling approach allows a detailed physical understanding of the phenomena that occur in a textile reinforced composite part. The integration approach of the project ensures a fluent transfer without critical data loss between the different stand-alone software tools. The project has set a first important step towards development of methods and standards. 1.8 Main project motivation Many projects show that textiles and textile reinforced composites have advantages in terms of cost and effort against prepreg technologies. The main differences which have been Page 8 of 36

9 observed are listed in table 3. Moreover, when it comes to the analysis, there is no universal approach to handle of textile reinforced composites as summarised in table 4. It can hence be concluded that there is a need for a new and integrated set of tool that enable a detailed analysis of textiles and textile reinforced composites. Table 3: motivation for textile preforming Prepreg resin pre-impregnated unidirectional tapes or fabrics (high material cost) Preform cost-effective dry fabrics (braids, weaves, NCF) storage limited by resin (6 months at 18 C) storage limited by fibre sizing (2 years at ambient conditions) labour-intensive stacking of single layers with defined orientation 2D fibre architecture Autoclave processing assembly by stitching or binders (high potential for automation even for complex shapes) option for 3D reinforcement by stitching reliable and cost-effective infiltration with non-autoclave techniques or RTM Table 4: Analysis options for composite materials Prepreg 2D fibre architecture nearly ideal fibre alignment micro / macro models established for determination of stiffness (unit cell models, CLT) strength (failure criteria WWFE) continuum damage evolution numerical implementation mainly for layered shell elements Preform 3D fibre architecture fibre misalignment due to draping + influence of infiltration process no universal (numerical) modelling approach, only single solutions structure and properties of the composite material are dependent on the manufacturing process simulation along the process chain based on 3M mechanics (micro / meso / macro) Page 9 of 36

10 2 Work performed and results The summary of the performed work and the gained results during the project follows the work package and task structure defined at the project start and was described in annex I of the consortium agreement. Due to a six month extension of the project this annex was updated including a reschedule of several deliverables and milestones. The following sections provide an overview of the planned work, aims and objectives within the different work packages and discusses the obtained results. 2.1 WP1: Material Characterisation / DataTool The aim of this work package is to collect all the information that may be relevant for any simulation within the framework. Related to this task is the creation of a comprehensive database (DataTool) to store the obtained data. The characterisation of different types of materials and material properties is essential and includes mechanical, physical and geometrical properties for both dry textiles and infiltrated composites. Especially for new textile architectures it is important to have an appropriate model which is able describe the material. Therefore work package 1 was subdivided in three logical sub-tasks: Task 1.1: Selection of relevant data The objective of task 1.1 was to identify a set of relevant parameters which can be used within any kind of simulation which is related to textile and textile composites. A comprehensive study on different textile architectures and their corresponding composite was performed to identify the key parameters on mechanical, physical and geometrical level. Moreover parameters are identified on micro, meso and macro-scales. From this study a list is compiled which contains the relevant material parameters. For each of these parameters a testing method and a short test description are given. Approximately 50 relevant parameters have been identified from the investigations in this task. These parameters are mainly geometrical properties of yarns and textile structures, mechanical properties of yarns, textiles and textile composites as well as deformability and permeability characteristics of dry textile. The identified parameters are listed in deliverable D Task 1.2: Structure of the DataTool Within the second task in work package 1, a structure for a database is developed. This database, which is named DataTool, will enable the storage of all relevant data identified in task 1.1 and will enable the different partners (or future users) to access this information. The structure of the DataTool does not only provide a way to store most important characteristics of a material, but also allows to add comments, link entries of different constituents or link the resulting property with its testing protocol. The results of this task, described in deliverable D1.2.1, are used within work package 5 where the DataTool is developed and implemented in a software application Task 1.3: Collection of data The third and final task of work package is the selection of relevant data. In this context the word relevant refers to the industrial validation examples which are defined in work package 6 (see section 2.6). Only for the selected materials used in the three validation parts and for Page 10 of 36

11 the selected process or external behaviour of the part, the required material parameters are obtained. The determination of the relevant parameters is done in three steps. First of all, a list of relevant data available from previous projects is compiled. This list contains references to data that may be relevant for the project. Second, data on experimental investigations on micro and meso scales are collected. This data is mainly used as input for the micro and meso-mechanical simulations by different partners. Finally, the data of experimental investigations on macro level are collected. This data is used for the validation of different simulations. All of the obtained values are finally listed three deliverables and are stored in the DataTool. Shear force normalised [N/mm] 1,6 1,4 1,2 1 0,8 0,6 0,4 0, Shear Angle ( ) Figure 3: examples of collected data within task WP2: 3M Mechanics The aim of work package 2 is to develop geometrical models of textile reinforcements on the one hand and mechanical models of textile composites on different levels on the other hand. Using both the finite element method and approximate methods (e.g. method of inclusions) the local stiffness and damage information of materials is in this way predicted. Three different structural levels are defined on which the models are defined: micro, meso and macro level. On micro (scale of 10 to 100 µm) level the individual fibres inside yarns or fibrous plies of the textile reinforcement are considered. The meso (scale 0.1 to 100 mm) level defines so-called unit cell of textile composites. Finally the macro (0.1 to 10 m) level defines the structural analysis level of a composite part. The models developed and used in this work package constitute the core of the integrated design tool. They provide crucial information of variability of mechanical properties of the composite material over a part, accounting for anisotropy of the properties and damage initiation/propagation. Work package 2 provides models compatible with and ready to be integrated in structural analysis tools, developed on Macro level in WP5. It also contributes to the description of the textile geometry, to the development of textile process models (WP3) and to the stiffness/failure/impact models (WP4). It uses the materials characterisation data provided in DataTool in WP1. More details about the work content and the results are summarised in the following task descriptions Task 2.1: Geometrical and mechanical models on Micro level In this task the uneven fibre distribution in yarns and fibrous plies is characterised from experimental observations and generalised in geometrical models of placement of fibres Page 11 of 36

12 accounting for irregularities of their distribution. The proposed models are implemented in the WiseTex software (figure 4). In case of yarns or tows it is found that the fibrous content usually decreases towards the edge of a bundle/yarn. The proposed geometrical model of the uneven fibrous content allows accounting for the major features of the experimentally observed microstructure of multifilament stitching yarns or fibrous bundles in woven fabrics. The experimental investigation of stitching sites in non-crimped fabrics shows a considerable distortion of fibres and a significant decrease of the fibrous content near the stitching sites. The failure and damage behaviour of fibre reinforced composites on micro-level is also investigated within this task. Parameters for failure criteria which are able to take into account three-dimensional loading as well as parameters for the continuum damage mechanics model of Ladevèze are characterised. The different damage and failure models are implemented in subroutines for finite element software. Figure 4: Graphical user interface of the new WiseTex modules for uneven fibre distributions in yarns (left) and fibrous plies (right) Task 2.2: Geometrical models on Meso level Geometrical models for both 3d-braids and structural stitches are developed within this task. Samples of both textile architectures are manufactured and used for geometrical and mechanical characterisation. Based on geometrical measurements, models using the WiseTex description are developed (figure 5). Page 12 of 36

13 Figure 5: WiseTex geometrical models of a 3d-braid (left) and structural stitching (right) Task 2.3: Homogenisation: Micro Meso level: Based on the damage models developed and implemented within task 2.1, the prediction of damage and decrease of mechanical properties associated to the damage are developed for meso-level. The obtained homogenised mechanical behaviour is compared to the experimental investigations. Initial damage events are predicted at approximately the same level as observed with acoustic emission measurements. The non-linear stress-strain behaviour of simple test coupons, the strength and ultimate strain are other quantities that are compared for the purpose of validating the model. Moreover microscopic computer tomography investigations of samples at different load levels are used for the validation of the damage pattern Task 2.4: Homogenisation: Meso Macro level In task 2.4 the homogenisation approach in case of high non-uniformity of internal structure and of stress-strain field is investigated. For multiply composites, a unit cell of one ply is often considered as a representative volume element (due to the computational restrictions). The classical analysis assumes infinite mapping of the unit cell in the out-of-plane direction. However, the meso strain field can be very much perturbed by the free surface. Several evidences are found about the preferential role of the surface layers for the damage accumulation of carbon-epoxy textile composite. The role of the free surface is explored numerically on a representative test problem and advanced homogenisation approaches are developed to take into account the above mentioned problems. The development of computational and modelling tools is also finalised in this task. A textile perform undergoes shear deformations when shaped into a 3D part. These deformations vary from point to point, changing the local properties of the composite part. The theoretical methods, implemented in software packages (WiseTex), allow calculation of the local stiffness in relation to the local deformation of performs, using meso-level description of the geometry of the unit cell of the reinforcement. The local deformation is predicted via simulation (QUIKFORM), and used together with the output of TexComp in FE packages (SYSPLY) as a material property data. The integrated model is implemented in a new version of SYSPLY software. It allows calculating for composites reinforced by woven or braided fabrics. 2.3 WP3: Process Simulation Textile raw materials (yarns, tows, roving) lose in strength and stiffness due to friction and bending during processing and manufacturing. The change of these properties influence the draping behaviour of this textile and the mechanical performance of a reinforced part manufactured with the textile. Work package 3 focuses on the simulation of the textile production to get the key information about a pre-cured textile structure (fibre orientations, fibre material properties including fibre-fibre friction). Material data characterised in WP1 and geometrical models developed in WP2 provide input to build models containing this key information. The models are used for further processing. Two important applications where the model is used are selected: the draping behaviour of a textile structure over a predefined geometry and the impregnation behaviour of a resin into a textile reinforced part. More details about the work content and the results are summarised in the following task descriptions. Page 13 of 36

14 2.3.1 Task 3.1: Development of concepts for a process simulation for NCF, sewing/stitching, 2D-braiding Task 3.1 is split into three parts. First, an extensive analysis of textile and textile composite manufacturing processes is performed resulting in a list of key parameters and machine settings which have an influence on the textile material properties. For multi-axial multi-play fabrics, the shear behaviour is analysed. The influence of the stitching process on the deformability is investigated for several textiles. Second, a virtual manufacturing model is developed that takes into account the observed phenomena in the first task. As a result a virtual textile is available including all necessary information (fibre orientations, fibre material properties, fibre-fibre friction) required for further use in other tasks. The virtual textile model is a meso-level model based on the meso-model of WP2. The third and final part of task 3.1 consists of validation of the developed models by compassion of experimental and simulated results Task 3.2: Drapability simulation Material laws base on experimental observation are developed to represent the deformability behaviour of fabrics. In a semi-discrete approach which is used, the necessary data appears to be the biaxial tensile surfaces and the in plane shear curve. Both of these properties can be determined experimentally of by simulation of the textile deformability (Figure 6) The developed material models are tested on a benchmark case which forces the draped fabric to significantly deform. Good agreement with experimental observations was obtained. Furthermore the data on the internal structure of the deformed fabric which is deaped on the benchmark part is used for injection simulation in task 3.3. Figure 6: A plain weave textile in undeformed, 27 sheared and 54 sheared deformation Task 3.3: Mould filling simulation The objective of this sub work package is to predict the flow of resin while it is infiltrating a textile reinforced part in the mould. Deformation of a textile structure can have a big influence on the permeability hence on the flow behaviour of the injection. A coupling between the draping simulation described in task 3.2 and this task is established. The finite particle method (FPM) is adapted to perform the injection simulation as it provides solutions for compressible and incompressible flow problems and can be coupled to structural FE codes. An important feature regarding polymer composites manufacturing science is that chemical reactions, heat transfer, temperature dependent viscosity can be handled. To illustrate the potential of the code, the case of numerical permeability prediction of a fabric unit cell has been simulated and compared with existing experimental data. Page 14 of 36

15 Figure 7: Draping simulation (left) and experiment(right) of a academic validation example. Figure 8: Comparison between the results of an infiltration simulation without (left) and with (right) taking into account draping results (figure 7). 2.4 WP4: External loading behaviour This work package looks at the complete structure of textile reinforced parts and aims to integrate macroscopic structural deformation, stress and failure modelling into. Global analysis methods which compute structural behaviour under external loads will be provided. The developed tools regard static stress, quasi-static failure, crash and dynamic impact computations. The input properties for the calculations in this work package are extracted from the results of simulations in work package 2 and 3. From the draping simulation performed in work package 3, the fibre orientations are extracted and used for the definition of the mechanical properties. The mechanical properties are determined by meso-scale calculations of a textile reinforced unit cell and homogenisation algorithms. More details about the work content and the results are summarised in the following task descriptions. Page 15 of 36

16 2.4.1 Task 4.1: Stress analysis and quasi-static failure criteria In this task the stress and failure behaviour of tri-axial braids as well as unstitched and through thickness thick NCF reinforced composites is investigated. Multi-level models which are able to predict damage inside the material are set up for the tri-axial braids. The mechanical behaviour of a part manufactured with this material is simulated and tested against experimental investigations. A good comparison could be observed. A special test rig, based on the Arcan test rig is developed to test through thickness properties of thick laminated structures. This test device is used to test unstitched and through thickness NCF reinforced composites. A failure envelope for out-off plain loading is deduced from the results and is used in macro-level part analysis Task 4.2: Impact and crash simulation In task 4.2 the dynamic part of the external loading behaviour is investigated. Damage development under low to high strain rates is investigated experimentally (figure 9) and tool for simulation of the observed behaviour are implemented and validated. Both unstitched and 3D reinforced materials hare investigated. The obtained material properties are used to simulate the bird impact behaviour of one of the industrial validation examples. Here delamination as well as in-plain failure modes are the focus of the work. Figure 9: Comparison between constitutive laws at low and high strain rates for fibre (left), transverse (middle) and shear (right) loading Task 4.3: Assistance for WP6 activities The relative new nature of the developments would require the industrial validation partners to study and learn these tools for the validation of the tool chain. This task was planned to provide support to the work package 6 activities and hereby reduce the time required to study the new tools. Manufacturing, testing, modelling and simulation activities are performed for each industrial validation partner. 2.5 WP5: Integration The main concept of focuses on an integrated solution. The integration of the different tools developed in the other work packages or from previous projects is handled in this work package. For a typical analysis of a textile reinforced composite part, two types of data can be determined. Already at the beginning of the proposal it was decided to separate these two types of data and develop different protocols to store them. The case depended data is data which will change for each new part that is designed. Examples are part geometry and mesh or stacking sequences. This data is stored in Page 16 of 36

17 generalised data transfer protocols which enable the different tools to exchange this data in this format. Second there is the case independent data which can be reused for different designs (e.g. material properties). The so called DataTool is developed to store this data in a database like environment. Case Caseindependent data Casedependent data Generalized data transfer protocols DataTool Datatool Components of WP1 WP3 GUI Components of WP4 WP3 Other Cases Components of WP2 WP3 Components of WP3 Figure Schematic overview of the DataTool and the general data transfer protocol More details about the work content and the results are summarised in the following task descriptions Task 5.1: Evaluation of the different tools A detailed analysis of which data is or will be transferred between the different software tools is performed. Typically each tool that is to be used is analysed by determining all possible inputs an outputs and identifying possible interactions with other tools. As a result not only a list of parameters is compiled but also the data type and the data flow is mapped. The result of this investigation is used for the definition of the data transfer protocols of the casedependent data and the DataTool for the case independent data Task 5.2: Data handling The data transfer protocols (DTP) are carefully designed, using input from WP 5.1, to be able to transfer case dependent data between different tools. The DTP developments are based on the XML standard and are able to transfer the definition of a mesh (nodes, elements, edges and faces), load and boundary condition descriptions as well as results (fields). Different steps are defined in the DTP: 1. Mapping: This step enables the mapping of data between different meshes. For instance the results from a draping simulation may be stored in a highly distorted mesh whereas a good quality mesh is required for the later mechanical analisis. 2. Transfer: This steps enables the transfer of data between nodal and integration point quantities. Page 17 of 36

18 3. Variable scaling: This step enables the modification of units as different partners may be using other unit systems. A second development within this task is a data format (DataTool) for storing the case independent data (typically material data). Also here it as chosen to develop a file format based on the XML standard. The data is divided in 1. Constituent: basic material used as an element to a global set. This can include epoxy resin, foam, Yarn. 2. Reinforcement: fibrous part in which the filaments are arranged in a particular way to achieve the desired result. This can include braided, stitched, woven fabrics, mats, rovings, UD. 3. Ply: Materials obtain by combination between reinforcement and a constituent. This can include UD fibre and epoxy resin, ceramic fibre and metallic matrix component of a laminate. 4. Laminate: A product made by bonding together ply or constituent like foam or honeycomb in order to form a single part Task 5.3: Design of the DataTool and a graphical user interface for A graphical user interface for both the data transfer protocols (figure 10 and figure 11) and the DataTool (figure 12). These tools enable the future users to easily handle the two developments of task 5.2 and ensure data integrity. Within the project mainly the features that are required for the industrial validation examples are supported. However, the tools are developed in such a way that they are expandable and can be easily enriched with new features. Figure 10: Left: A graphical representation of the steps in the data transfer protocols. Right: Page 18 of 36

19 graphical user interface of the mapping step in the data transfer protocols. Figure 11: Left: graphical user interface of the transfer step in the data transfer protocols. Right: graphical user interface of the variable scaling step in the data transfer protocols. Figure 12: The graphical user interface of the DataTool 2.6 WP6: Industrial validation / Test This WP6 intends to provide test cases for the "industrial" acceptance of the integrated simulation tool developed in the other work packages. The goal is to test the whole simulation chain, including the simulation of the textile production (creation of the "virtual" textile), the prediction of meso/macro properties, the drapability behaviour, the impregnation and FE-analysis of the finished composite part. A comparison between the manufacturing (including testing) and the whole process and mechanical simulation will be performed. The main technologies of textile production as braiding, stitched material including non crimpfabrics and fibre placement will be investigated. The goal of the defined test cases is to become the basis of an "industrial" acceptance methodology for future improvements or upgrades of the developed simulation tool. Page 19 of 36

20 More details about the work content and the results are summarised in the following task descriptions Task 6.1: Specification of test cases In order to test different developments of the project, the industrial partners have selected complementary examples. As soon as the project started the validation examples are defined and for each of them the following features are defined: a description of the different parts to be manufactured (material, technology, structure ), the kind of mechanical tests to be lead on the defined parts and the tool chain to be performed for each part (both process and mechanical simulations). A graphical representation of the three validation examples is shown in figure 13. Figure 13: Graphical representations of the three industrial validation examples Task 6.2: Manufacturing of the parts and testing Based on the information coming from 6.1 the activities in this task vary for each validation example. Basically three kinds of activities can be distinguished. First, a set of basic material tests each material used in the validation example is performed. This task is required to characterise the materials used in the validation examples. Second, the industrial validation parts are manufactured and different properties are determined during or after the manufacturing. This includes infiltration, draping and mechanical characteristics. Finally the mechanical performance of the parts is determined. Buckling and bird impact behaviour are investigated here. Figure 14: Manufacturing steps for the EADS-G validation example In this task, the industrial partners are supported by other partners as previously described in section Page 20 of 36

21 2.6.3 Task 6.3: Tool chain proof on defined test cases The aim of task 6.3 is the testing of the whole simulation chain and validating it against experimental results. For each industrial partner the chain is validated on three levels: First, simulations on the basic material test enable to validate developments issued from work package 2, 3 and 4. Second, a feature of the part manufacturing process is simulated and validated against its experimental counterpart. Finally, a validation of the mechanical performance prediction is validated. EADS-G validation part: Internal geometry CAD Model & Mesh generation WiseTex model geometrical WP3 Draping model Shear angle measurements COMPARE Mechanical test on specimens 0,9 0,8 Calculated mechanical properties COMPARE 0,7 0,6 0,5 0,4 0,3 0,2 0, ,01 0,02 0,03 0,04 0,05 0,06-0,1 Bird strike Bird strike COMPARE simulation testing Figure 15: Graphical representations of the three industrial validation examples Task 6.4: Cost benefit analysis To evaluate the benefit of the developments against up to now standard approaches, a cost-benefit analysis is performed. The analysis is based on the industrial validation examples and is performed on cost and effort levels. For all three validation examples the approach has shown a benefit in both cost (between 4% and 60%) end effort (between 15% and 70%). Page 21 of 36

22 80% 70% ALA - approach DAS - approach EADS - approach 68,7% 64,1% 59,5% 60% 50% 40% 30% 20% 10% 24,2% 15,4% 14,2% 19,2% 9,6% 3,9% 0% Cost benefit Effort benefit Total benefit Figure 16: results of the cost benefit analysis for the three industrial validation examples. 2.7 WP7: Design guidelines / Standardisation The optimum design and cost effective manufacturing of textile structural composites requires a lot of knowledge on the factors of influence. At the early phase of development the designer has to know which basic textile process is the best for the specific structure, which limitations have to be taken into consideration and which detailed parameters lead to an optimum weight saving. Work package 7 tries to generate design guidelines, design diagrams and rules of thumb, based on the work done within the project and are intended to help future engineers to design with textile reinforced composite materials. More details about the work content and the results are summarised in the following task descriptions Task 7.1: Definition of structural elements In order to structure the magnitude of information, characteristic sub-structures that can be manufactured using textile reinforced composites are identified. The sub-structures identifies are: shells and panels (flat, 2d curved or 3d curved), profiles, box structures, interfaces and joints, overlapping linear joints (riveted or bonded), fittings and cut-outs. Figure 17: a selection of the identified structural elements. From left to right: panel with jointed profiles, cut-out, profile, joint element. Page 22 of 36

23 All characteristic elements that are used in the industrial validation examples are included for the further investigation in work package Task 7.2: Potential and limitations of textile processes The state-of-the-art for the textile processes are investigated with the aid of literature studies and own experience of partners. The following processes, which are used in the industrial validation examples, are investigated: non-crimp fabrics, robot assisted braiding, stitching and textile fibre placement (embroidery). For each of them the possibilities and limitations are listed and described in deliverable D Task 7.3: Development of design guidelines The goal of task 7.3 is to show procedures how to design the basic elements which were proposed in task 7.1 with the processes described in task 7.2. The basis for the determination of optimum design is the use of optimisation techniques with gradient and generic algorithms. From the results of the optimisation calculations engineering design criteria and rules of thumb are deduced and summarised in deliverable D Task 7.4: Evaluation based on realistic applications The partners which manufactured the industrial validation parts tried to evaluate the design guidelines fro task 7.3 based on their part. The applicability of the rules is tested and remarks/comments are collected and described in deliverable D Task 7.5: Standardization Content of this work-package is to derive and generate the basis and philosophies for defining standards for test procedures and textile processes. Since textile composites offer a high degree of freedom in respect to fibre architectures and production methods the effort for material testing, accreditation, quality control and simulation is high for this kind of materials. Based on the experience within the project and the evaluation of the produced tools and parts standards are proposed within this task. Page 23 of 36

24 2.8 List of deliverables The list of deliverables is given in Error! Reference source not found.. Table 5: Deliverables list WP1 Del. no Deliverable name Lead contractor D1.1.1 Relevant Data ITA D1.2.1 Requirements for DataTool ITA D1.3.1 List of available data ITA D1.3.2 Data at mesoscopic and microscopic scale ITA D1.3.3 Complete data ITA D2.1.1 Model of fibre placement on micro-level KUL D2.1.2 Model of damage on micro-level with parameters identified from experiments DLR D2.2.1 Geometrical model of architecture of 3d braids implemented in FEA and WiseTex ZARA D2.2.2 Geometrical model and architecture of structural stitching implemented in WiseTex KUL D2.2.3 Validation of the geometrical models of 3D braids ZARA D2.2.4 Validation of the geometrical models of structural stitching KUL D2.3.1 Model of damage initiation and development on meso-level KUL D2.3.2 Validation of damage models on meso-level EADS-G D2.4.1 Theoretical formulation of homogenisation on meso-level for high nonuniformity of stress-strain fields KUL D2.4.2 Validation of the meso-mechanical stiffness models for high non-uniformity of stress-strain fields KUL D2.4.3 Computational tools implementing the 3M models and integrated with Macro-analysis KUL D3.1.1 Models of deformability of non-crimp fabrics, of the stitched reinforcements and of internal geometry of deformed stitched reinforcement KUL D3.1.2 Definition of the influence of different machine settings on the characteristics of a textile ITA D3.1.3 Geometrical description of the positions of fibres and stitches of different textiles including open and closed structures. ITA D3.2.1 Identification of meso-macro mechanical behaviour laws used in the fabric finite element approach developed for fabric forming simulations. LAMCOS D3.2.2 Specific meso-macro finite element for dry textile forming. D3.2.3 Report including modelling and validation of dry fabrics forming. LAMCOS EADS-F D3.3.1 First version of a numerical model for the prediction of permeability ESI D3.3.2 Numerical model of prediction of permeability ESI D3.3.3 Injection modelling and validation of deformed dry fabrics. D4.1.1 Final version of software tools for quasi static failure EADS-F DLR Page 24 of 36

25 Deliverable name Lead contractor Final report on validation of homogenisation EADS-G D4.1.3 Final report on experimental validation tests and FE stress and failure analysis DLR D4.2.1 Preliminary version of software tool for dynamic failure analysis CRAN D4.2.2 Final version and final technical report on validation of software tool for dynamic failure analysis CRAN D4.2.3 Final technical report on crash simulation and failure studies SISPRA D4.3.1 Final report on supportive activities to WP6 D5.1.1 Analysis of the data used and produced during a composite analysis DLR ESI D5.2.1 Data and transfer protocol ESI D5.3.1 GUI D6.1.1 D6.2.1 Test case definition ESI DAS Coupon testing DAS D6.2.2 Manufactured and tested industrial parts DAS D6.3.1 tool chain proof DAS D6.4.1 Cost benefit analysis D7.1.1 D7.2.1 Basic elements requirements DAS IFB Performance catalogue IFB D7.3.1 Design guidelines IFB D7.3.2 Basis for handbook chapter IFB D7.4.1 Guideline Evaluation IFB D7.5.1 Standardization philosophies IFB Del. no D4.1.2 Page 25 of 36

26 3 Degree to which the objectives where reached Objective: reliable simulation tools and design principles Within the project it was chosen to use integrate existing and newly developed tools which have been extensively tested against experimental benchmark cases. Also within the project a large amount of effort was put into experimental work for validation purposes. Objective: simulation along the process line with a virtual manufacturing chain. Taking into account the prediction of textile deformations during draping processes and linking the results from there simulations to predictions or permeability and mechanical properties a simulation along the process line is established. Moreover the tools are organised in such a way that other features that have not been taken into account during the project can easily be integrated. Objective: development of an adequate integrated simulation tool The integration of the existing and newly developed tools have led to reliable simulation tools and design principles which are able to handle composites reinforced with 3d fibre architectures. The linking and integration of exiting tools ensures a fluent interaction and data interchange with minimum friction and without critical data loss. Furthermore a graphical user interface is developed to minimise user interference. Objective: consider braiding, weaving and stitching technologies The performance of textiles and textile reinforced composites is analysed on three different approximation levels called 3M (micro / meso / macro) mechanics. This approach enables to consider many different kinds of textile architectures as all of them can be assembled from the basic building blocks. Newly developed tools enable geometrical and mechanical descriptions of braiding, advanced engineering textiles, weaving and stitching textile preforming technologies. Objective: improve properties in the thickness direction with stitching technologies Different though thickness reinforcement technologies where investigated. New modelling tools are developed that enable the analysis of such materials. Via optimisation algorithms ideal stitching configurations could be determined. These results where experimentally validated. Objective: proof for this integration concept and benefit of cost and effort The integration concept is validated for different application fields of textile preformed composites in Aerospace. Typical stiffened skin sections with integral joining technologies are evaluated. The use of the approach has shown to provide a reduction of in 4% to 60% of costs and a reduction of 15% to 70% of effort for the industrial validation examples. Objective: extraction of design rules for the use of textiles as composite reinforcement Physical understanding of textile preformed composites behaviour is obtained during the project and is summarised in design rules. Many of the obtained results have been disseminated on conferences and in international journals. Page 26 of 36

27 4 Achievements of the project related to the state of the art 4.1 State of the art before project start Most preforming technologies are well-known in the textile industry and have reached a high level of automation. By this, a cost-effective manufacturing of fabrics and apparel has been established. The idea of adapting the technologies of braiding, weaving and stitching for reinforcement of composites started in the 80 s and was improved mainly in the late 90 s. The problems to be solved have been fibre damage, due to the stiff and brittle fibre behaviour, friction, fibre waviness, drapability and infiltration and the minor decrease of in-plane mechanical properties of textile preformed composites in comparison with unidirectional prepregs. Meanwhile dry fibre textiles have passed the prototype stage and are available for application. Many of the basics have been developed in National and CEC funded research projects such as INTEX, MULTEXCOMP or COBRAID. The main results of the related research activities are that: textile structural composites have proven their high potential for automation and cost reduction in the manufacturing of high performance composites, 3d-reinforcements, especially realized by stitching, have proven a high potential for improve damage tolerance, structural integrity and energy absorption without significant reducing in-plane performance, several series applications are in development making use of stitched textile preforms (for example Airbus A380 pressure dome, Boeing wing-structures and Mercedes SLR crash structures) Figure 4.1.1: Circular braider at EADS-G Figure 4.1.2: Stitching of a textile preform However, only basic principles for simulation and calculation as well as only limited understanding of principles and mechanics of these techniques are available. Most of the theories and finite element implementations for composite materials are based on simple unidirectional homogenisation methods of fibre and matrix properties; strength prediction is done only plywise. For 3d fibre architectures homogenisation of voxel-type representative volume elements (RVE) has shown promising results, whilst failure prediction is much more Page 27 of 36

28 difficult due to nonlinearities like friction behaviour of cross-over points, complex 3d fibrematrix interactions. For parts of the regarded simulation chain, including manufacturing and loading stage, stand-alone solutions are already available resp. in development phase, e.g. PAMFORM or WiseTex. But, up to now, communication and data handling between these tools is not harmonised so that a fluent exchange of data is not possible. Figure 4.1.3: Simulation of fibre reorientation due to draping process Figure 4.1.4: Modelling of textile preforms with WiseTex software Furthermore, the know-how on manufacturing (draping, tooling and impregnation) of integrated, complex-shaped preforms is only limited. This results in large numbers of product development iteration loops and, furthermore, in an extensive testing effort necessary for qualification of dry fibre textiles, limited always to one single configuration. The aim of is to eliminate the gaps between the potential in application and the understanding and analysis of textile preformed composites. To enable the design engineer to take advantage of these types of materials, practical guidelines like the HSB (Handbuch Strukturberechnung), widely used in Aerospace construction and analysis of metals and unidirectional reinforced prepregs, are further missing prerequisites to be worked out in the proposed project by collating available information in documents and educational tools. Page 28 of 36